was leading in the polls. the fact that it's not trending towards them now seems odd. megyn: the questions were raised because they chaingsd it so close teen election day, tucker. >> the cell phone change is not the only change that was made. the key change is in the sample, who gets the calls. they increased the size of the non-white sample in their sample set. megyn: that's one of the things axlerod wanted them to do. >> that's the key here. i have no knowledge that gallup is doing anything wrong or doing it because of pressure the federal government or the obama campaign. i'm merely noting this is a big deal. it has had a measurable change on the poll outcome on the band wagon effect is real. if people look at the polls and think somebody is going to win, if it's clear that person is leading, that person tends to get more support because he is leading. this is not a small thing. >> i would say we are talking about fractions and 1%, 2%. these aren't big massive jumps. truly the methodologies are important be